Implementing the paper "Facemask Wearing Alert System Based on Simple Architecture With Low-Computing Devices" using a two-step paradigm, detected faces with Single Shot Detector, and classified into two classes (with/without mask) using a Keras convolutional neural network.
You should first clone this project:
git clone https://github.com/ZahraDehghani99/FacemaskWearingAlertSystem.git
cd FacemaskWearingAlertSystem
Then you should create a conda environmet as follows:
conda create -n facemask_classification python==3.7 pip==20.2.4
After creating a conda environment, you should activate it and then run the following code to install requirements.
pip install -r requirements.txt
After installing requirements, you should run detect_mask_image.py as follows:
python detect_mask_image.py --image examples/1.png
Note : you should change address of face_detector, mask_detector base on your local address.
After run this command if the system predicts without_mask, the FaceMask_detection_alert will be broadcast.
Note : In the MTCC_face_detector I tried face detection using MTCNN.